These analyses reveal a complex relationship between crime rates and economic factors, racial composition and age, suggesting that crime is not just a single social problem, but is inffuenced by the interaction of multiple factors. An in-depth analysis of the association between a number of social factors and crime rates in the United States reveals that areas with higher per capita incomes typically have lower crime rates, while areas with higher levels of poverty and lower levels of education tend to face higher crime pressures. This phenomenon may be closely related to the scarcity of resources, higher unemployment and the existence of social inequalities.
In addition, racial composition plays a signiffcant role in crime rates. There are differences in the extent to which different racial groups are involved in different types of crime, and certain racial groups may be exposed to a higher risk of crime in particular neighborhoods.
These analyses provide a more complete understanding of the spatial distribution of crime rates across U.S. states and the multiple factors that inffuence this distribution. This understanding provides policymakers with a more speciffc social context to help them design more targeted and effective interventions to reduce crime rates.
---
title: "Analysis of 2018 Crime Data in the United States"
output:
flexdashboard::flex_dashboard:
orientation: rows
vertical_layout: scroll
fig_width: 8
fig_height: 7
source_code: embed
---
```{r setup, include=FALSE}
library(flexdashboard)
library(tidyverse)
library(ggplot2)
library(dplyr)
library(DT)
library(knitr)
library(rpivotTable)
library(plotly)
library(openintro)
library(highcharter)
library(ggvis)
library(shiny)
library(viridis)
library(RColorBrewer)
library(corrplot)
library(ggpubr)
library(lmtest)
library(tidyr)
library(corrplot)
library(maps)
library(highcharter)
library(heatmaply)
library(scales)
```
HeatMap
========================================
```{r}
# Data loading and crime related variables
crime_data <- read.csv("processed_crimedata.csv")
crime_related_vars <- c("murders", "murdPerPop", "rapes", "rapesPerPop", "robberies", "robbbPerPop",
"assaults", "assaultPerPop", "burglaries", "burglPerPop", "larcenies",
"larcPerPop", "autoTheft", "autoTheftPerPop", "arsons", "arsonsPerPop",
"ViolentCrimesPerPop", "nonViolPerPop")
# Ensure variable order, grouping
all_variables <- c("population", "householdsize", "racepctblack", "racePctWhite", "racePctAsian",
"racePctHisp", "agePct12t21", "agePct12t29", "agePct16t24", "agePct65up",
"numbUrban", "pctUrban", "medIncome", "pctWWage", "pctWFarmSelf", "pctWInvInc",
"pctWSocSec", "pctWPubAsst", "pctWRetire", "medFamInc", "perCapInc", "whitePerCap",
"blackPerCap", "indianPerCap", "AsianPerCap", "OtherPerCap", "HispPerCap",
"NumUnderPov", "PctPopUnderPov", "PctLess9thGrade", "PctNotHSGrad", "PctBSorMore",
"PctUnemployed", "PctEmploy", "PctEmplManu", "PctEmplProfServ", "PctOccupManu",
"PctOccupMgmtProf", "MalePctDivorce", "MalePctNevMarr", "FemalePctDiv",
"TotalPctDiv", "PersPerFam", "PctFam2Par", "PctKids2Par", "PctYoungKids2Par",
"PctTeen2Par", "PctWorkMomYoungKids", "PctWorkMom", "NumKidsBornNeverMar",
"PctKidsBornNeverMar", "NumImmig", "PctImmigRecent", "PctImmigRec5", "PctImmigRec8",
"PctImmigRec10", "PctRecentImmig", "PctRecImmig5", "PctRecImmig8", "PctRecImmig10",
"PctSpeakEnglOnly", "PctNotSpeakEnglWell", "PctLargHouseFam", "PctLargHouseOccup",
"PersPerOccupHous", "PersPerOwnOccHous", "PersPerRentOccHous", "PctPersOwnOccup",
"PctPersDenseHous", "PctHousLess3BR", "MedNumBR", "HousVacant", "PctHousOccup",
"PctHousOwnOcc", "PctVacantBoarded", "PctVacMore6Mos", "MedYrHousBuilt",
"PctHousNoPhone", "PctWOFullPlumb", "OwnOccLowQuart", "OwnOccMedVal", "OwnOccHiQuart",
"OwnOccQrange", "RentLowQ", "RentMedian", "RentHighQ", "RentQrange", "MedRent",
"MedRentPctHousInc", "MedOwnCostPctInc", "MedOwnCostPctIncNoMtg", "NumInShelters",
"NumStreet", "PctForeignBorn", "PctBornSameState", "PctSameHouse85",
"PctSameCity85", "PctSameState85", "LandArea", "PopDens", "PctUsePubTrans",
"LemasPctOfficDrugUn")
# Group
available_columns <- colnames(crime_data)
groups <- list()
for (i in seq(1, length(all_variables), by = 30)) {
end_idx <- min(i + 29, length(all_variables))
group <- unique(c(crime_related_vars, all_variables[i:end_idx]))
valid_group <- group[group %in% available_columns]
groups <- append(groups, list(valid_group))
}
```
Column
--------------------------------------------------
### Group1
```{r}
# Group 1
group1_data <- crime_data %>% select(all_of(groups[[1]]))
correlation_matrix1 <- cor(group1_data, use = "complete.obs")
diag(correlation_matrix1) <- NA
heatmaply(
correlation_matrix1,
main = "Heatmap for Group 1",
xlab = "Variables",
ylab = "Variables",
fontsize_row = 8,
fontsize_col = 8,
width = 750,
height = 600,
Rowv = FALSE,
Colv = FALSE
)
```
### Group2
```{r}
group2_data <- crime_data %>% select(all_of(groups[[2]]))
correlation_matrix2 <- cor(group2_data, use = "complete.obs")
diag(correlation_matrix2) <- NA
heatmaply(
correlation_matrix2,
main = "Heatmap for Group 2",
xlab = "Variables",
ylab = "Variables",
fontsize_row = 8,
fontsize_col = 8,
width = 750,
height = 600,
Rowv = FALSE,
Colv = FALSE
)
```
Column
--------------------------------------------------
### Group3
```{r}
group3_data <- crime_data %>% select(all_of(groups[[3]]))
correlation_matrix2 <- cor(group3_data, use = "complete.obs")
diag(correlation_matrix2) <- NA
heatmaply(
correlation_matrix2,
main = "Heatmap for Group 2",
xlab = "Variables",
ylab = "Variables",
fontsize_row = 8,
fontsize_col = 8,
width = 750,
height = 600,
Rowv = FALSE,
Colv = FALSE
)
```
### Group4
```{r}
group4_data <- crime_data %>% select(all_of(groups[[4]]))
correlation_matrix2 <- cor(group4_data, use = "complete.obs")
diag(correlation_matrix2) <- NA
heatmaply(
correlation_matrix2,
main = "Heatmap for Group 2",
xlab = "Variables",
ylab = "Variables",
fontsize_row = 8,
fontsize_col = 8,
width = 750,
height = 600,
Rowv = FALSE,
Colv = FALSE
)
```
Map + Pie Chart
====================================
```{r, include=FALSE}
# Load the necessary libraries
library(flexdashboard)
library(ggplot2)
library(dplyr)
library(maps)
library(plotly)
library(tidyr)
# Load the dataset
crimedata <- read.csv("crimedata.csv")
# Calculate the average crime rate and other statistics for each state
state_crime_rate <- crimedata %>%
group_by(state) %>%
summarise(
LandArea = mean(LandArea, na.rm = TRUE),
perCapInc = mean(perCapInc, na.rm = TRUE),
population = mean(population, na.rm = TRUE),
PopDens = mean(PopDens, na.rm = TRUE),
ViolentCrimesPerPop = mean(ViolentCrimesPerPop, na.rm = TRUE),
nonViolPerPop = mean(nonViolPerPop, na.rm = TRUE),
murders = mean(murders, na.rm = TRUE),
rapes = mean(rapes, na.rm = TRUE),
robberies = mean(robberies, na.rm = TRUE),
assaults = mean(assaults, na.rm = TRUE),
burglaries = mean(burglaries, na.rm = TRUE),
larcenies = mean(larcenies, na.rm = TRUE),
autoTheft = mean(autoTheft, na.rm = TRUE),
arsons = mean(arsons, na.rm = TRUE),
agePct12t21 = mean(agePct12t21, na.rm = TRUE),
racepctblack = mean(racepctblack, na.rm = TRUE)
)
# Fill missing values with the median
state_crime_rate <- state_crime_rate %>%
mutate(across(everything(), ~ replace_na(., median(., na.rm = TRUE))))
# Load US map data
us_map <- map_data("state")
# Merge crime data with map data
state_abbreviation <- data.frame(state = state.abb, region = tolower(state.name))
state_crime_rate <- state_crime_rate %>%
left_join(state_abbreviation, by = "state")
merged_data <- us_map %>%
left_join(state_crime_rate, by = "region")
```
Row
--------------------------------
### Crime Rate Map
```{r}
# Set predefined choices (no Shiny controls)
crime_type <- "murders"
age_group <- "agePct12t21"
race_group <- "racepctblack"
general_info <- "population"
# Prepare data for plotting
merged_data <- merged_data %>%
mutate(
crime_rate = .data[[crime_type]],
age_rate = .data[[age_group]],
race_rate = .data[[race_group]],
general_info = .data[[general_info]]
)
# Calculate state centers for labeling
state_centers <- merged_data %>%
group_by(region) %>%
summarize(
long = mean(long, na.rm = TRUE),
lat = mean(lat, na.rm = TRUE),
state = unique(state)
)
# Create the map
p <- ggplot(merged_data, aes(
x = long, y = lat, group = group,
fill = crime_rate,
text = paste(
"State:", region, "<br>",
"Crime Type (Murders):", round(crime_rate, 2), "<br>",
"Age Group (12-21%):", round(age_rate, 2), "<br>",
"Race Group (Black%):", round(race_rate, 2), "<br>",
"General Info (Population):", round(general_info, 2)
)
)) +
geom_polygon(color = "white") +
scale_fill_gradient(low = "#F2D0D0", high = "#C3272B", name = "Crime Rate") +
coord_fixed(ratio = 1.3) +
geom_text(data = state_centers, aes(x = long, y = lat, label = state),
color = "black", size = 3, fontface = "bold", inherit.aes = FALSE) +
theme_minimal() +
labs(
title = "Average Murders by State",
subtitle = "Hover to see state details",
x = "", y = ""
) +
theme(
axis.text = element_blank(),
axis.ticks = element_blank(),
panel.grid = element_blank()
)
# Convert to Plotly for interactivity
ggplotly(p, tooltip = "text", width = 1300, height=700)
```
Column {data-width=50%}
------------------------------------
```{r}
# Count the total number of different crime types
crime_data <- data.frame(
CrimeType = c("Murders", "Rapes", "Robberies", "Assaults",
"Burglaries", "Larcenies", "AutoTheft", "Arsons"),
Count = colSums(crimedata[, c("murders", "rapes", "robberies", "assaults",
"burglaries", "larcenies", "autoTheft", "arsons")], na.rm = TRUE)
)
# Normalized percentage
crime_data <- crime_data %>%
mutate(Percentage = Count / sum(Count) * 100)
# Adjust the percentages to make sure the sum is 100%
error <- 100 - sum(crime_data$Percentage) # Calculation error
crime_data$Percentage[which.max(crime_data$Percentage)] <-
crime_data$Percentage[which.max(crime_data$Percentage)] + error # Assign the error to the maximum value
# Draw the pie chart using plotly
crime_pie <- plot_ly(
crime_data, labels = ~CrimeType, values = ~Percentage, type = 'pie',
textinfo = 'label+percent',
insidetextorientation = 'auto',
marker = list(colors = c('#FF5733', '#FADA54', '#81CAC3', '#FFAE8C', '#FFA63B', '#35A9DA', '#D1AECE'))
) %>%
layout(
title = "Crime Data Distribution (1995)",
width = 700, height = 650,
margin = list(l = 0, r = 0, t = 50, b = 110)
)
crime_pie
```
```{r}
# Calculate average race data
race_data <- data.frame(
RaceGroup = c("Black", "White", "Asian", "Hispanic"),
Percentage = colMeans(crimedata[, c("racepctblack", "racePctWhite", "racePctAsian", "racePctHisp")], na.rm = TRUE)
)
# Race distribution pie chart
rp <- plot_ly(
race_data, labels = ~RaceGroup, values = ~Percentage, type = 'pie',
textinfo = 'label+percent',
insidetextorientation = 'radial',
marker = list(colors = c('#9999FF', '#FF9999', '#40c2a8', '#FFD700'))
) %>%
layout(
title = "Average Race Distribution (Percentage)",
width = 700, height = 650,
margin = list(l = 0, r = 0, t = 50, b = 110),
pie = list(domain = list(x = c(0.4, 0.5), y = c(0.4, 0.5)))
)
# Print the pie chart
rp
```
Column{data-width=50%}
------------------------------------
```{r}
# Detailed education level pie chart
detailed_education_data <- data.frame(
Category = c("Less than 9th Grade", "Not High School Graduate", "Bachelor's or More"),
Percentage = colMeans(crimedata[, c("PctLess9thGrade", "PctNotHSGrad", "PctBSorMore")], na.rm = TRUE)
)
# Detailed education level pie chart
dep <- plot_ly(
detailed_education_data, labels = ~Category, values = ~Percentage, type = 'pie',
textinfo = 'label+percent',
insidetextorientation = 'auto',
marker = list(colors = c('#C9EBD8', '#FFBADC', '#f7d323'))
) %>%
layout(
title = "Detailed Education Levels (Percentage)",
width = 700, height = 600,
margin = list(l = 0, r = 0, t = 50, b = 80),
pie = list(domain = list(x = c(0.4, 0.5), y = c(0.4, 0.5)))
)
# Print the pie chart
dep
```
```{r}
# Calculate average crime data
crime_data <- data.frame(
CrimeType = c("ViolentCrimesPerPop", "nonViolPerPop"),
AverageRate = colMeans(crimedata[, c("ViolentCrimesPerPop", "nonViolPerPop")], na.rm = TRUE)
)
# Create an interactive pie chart for the crime data
ep_crime <- plot_ly(
crime_data, labels = ~CrimeType, values = ~AverageRate, type = 'pie',
textinfo = 'label+percent',
insidetextorientation = 'radial',
marker = list(colors = c('#DC143C', '#00CED1'))
) %>%
layout(
title = "Average Crime Rates (Per 100,000 People)",
width = 700, height = 600,
margin = list(l = 0, r = 0, t = 50, b = 60),
pie = list(domain = list(x = c(0.4, 0.5), y = c(0.4, 0.5)))
)
# Print the crime pie chart
ep_crime
```
Total Crimes Ranking by States
=====================================
```{r include=FALSE}
crimedata <- read.csv("crimedata.csv")
selected_data <- crimedata %>%
select(state, murders, rapes, robberies, assaults, burglaries, larcenies, autoTheft, arsons)
selected_data <- selected_data %>%
mutate(
murders = ifelse(is.na(murders), median(murders, na.rm = TRUE), murders),
rapes = ifelse(is.na(rapes), median(rapes, na.rm = TRUE), rapes),
robberies = ifelse(is.na(robberies), median(robberies, na.rm = TRUE), robberies),
assaults = ifelse(is.na(assaults), median(assaults, na.rm = TRUE), assaults),
burglaries = ifelse(is.na(burglaries), median(burglaries, na.rm = TRUE), burglaries),
larcenies = ifelse(is.na(larcenies), median(larcenies, na.rm = TRUE), larcenies),
autoTheft = ifelse(is.na(autoTheft), median(autoTheft, na.rm = TRUE), autoTheft),
arsons = ifelse(is.na(arsons), median(arsons, na.rm = TRUE), arsons)
)
state_crime_totals <- selected_data %>%
group_by(state) %>%
summarise(
murders = sum(murders, na.rm = TRUE),
rapes = sum(rapes, na.rm = TRUE),
robberies = sum(robberies, na.rm = TRUE),
assaults = sum(assaults, na.rm = TRUE),
burglaries = sum(burglaries, na.rm = TRUE),
larcenies = sum(larcenies, na.rm = TRUE),
autoTheft = sum(autoTheft, na.rm = TRUE),
arsons = sum(arsons, na.rm = TRUE),
total_crimes = sum(murders, rapes, robberies, assaults, burglaries, larcenies, autoTheft, arsons, na.rm = TRUE) # Calculate the total
)
print(state_crime_totals)
```
Column {data-height=500}
-------------------------------------------
### Total Crimes Ranking by States
```{r}
p <- ggplot(state_crime_totals,
aes(x = reorder(state, -total_crimes), y = total_crimes,
fill = total_crimes,
text = paste('State: ', state, '<br>Total Crime: ', total_crimes))) +
geom_bar(stat = "identity") +
scale_fill_gradient(low = "#ADD8E6", high = "#8A2BE2") +
labs(title = "Total Crimes Ranking by States",
x = "States",
y = "Total Crimes",
fill = "Total Crimes"
) +
theme_minimal() +
theme(panel.grid.major = element_blank(),
panel.grid.minor = element_blank())
p_interactive <- ggplotly(p, tooltip = "text")
p_interactive
```
Column {data-height=500}
---------------------------------------
```{r include=FALSE}
long_data <- selected_data %>%
group_by(state) %>%
summarise(
murders = sum(murders, na.rm = TRUE),
rapes = sum(rapes, na.rm = TRUE),
robberies = sum(robberies, na.rm = TRUE),
assaults = sum(assaults, na.rm = TRUE),
burglaries = sum(burglaries, na.rm = TRUE),
larcenies = sum(larcenies, na.rm = TRUE),
autoTheft = sum(autoTheft, na.rm = TRUE),
arsons = sum(arsons, na.rm = TRUE)
) %>%
pivot_longer(
cols = -state,
names_to = "crime_type",
values_to = "crime_count"
) %>%
mutate(
crime_type = case_when(
crime_type %in% c("murders", "rapes", "arsons") ~ "Other",
TRUE ~ crime_type
)
) %>%
group_by(state, crime_type) %>%
summarise(
total_crime_count = sum(crime_count, na.rm = TRUE),
.groups = "drop"
) %>%
group_by(state) %>%
mutate(
total_crime_in_state = sum(total_crime_count, na.rm = TRUE),
crime_per = total_crime_count / total_crime_in_state
) %>%
arrange(state, desc(crime_per)) %>%
ungroup()
print(long_data)
```
### Crime Plot {.plotly}
```{r}
crime_type_order <- long_data %>%
group_by(crime_type) %>%
summarise(total_crime = sum(total_crime_count, na.rm = TRUE), .groups = "drop") %>%
arrange(desc(total_crime)) %>%
pull(crime_type)
#
top_states <- long_data %>%
group_by(state) %>%
summarise(total_crime = sum(total_crime_in_state, na.rm = TRUE), .groups = "drop") %>%
arrange(desc(total_crime)) %>%
slice_head(n = 5) %>%
pull(state)
# Sort by total crime count and select the top 5 states
filtered_data <- long_data %>%
filter(state %in% top_states) %>%
mutate(crime_type = factor(crime_type, levels = crime_type_order))
p <- ggplot(filtered_data, aes(
x = state,
y = crime_per,
fill = crime_type,
text = paste(
"State: ", state,
"<br>Crime Type: ", crime_type,
"<br>Proportion: ", percent(crime_per, accuracy = 0.1)
)
)) +
geom_bar(stat = "identity", position = "stack") +
geom_text(
aes(label = percent(crime_per, accuracy = 0.1)),
position = position_stack(vjust = 0.5),
size = 3,
color = "black"
) +
scale_y_continuous(labels = percent) +
scale_fill_brewer(palette = "Set3") +
theme_minimal() +
theme(
axis.text.x = element_text(angle = 45, hjust = 1),
axis.title.x = element_blank(),
axis.title.y = element_blank(),
legend.title = element_blank(),
plot.margin = margin(1, 2, 1, 1, "cm")
) +
labs(
title = "Crime Types in Top 5 States",
fill = "Crime Type"
)
# Convert to an interactive chart
p_interactive <- ggplotly(p, tooltip = "text")
p_interactive
```
Column 1 {.column data-width=600}
-----------------------------------------------------------------------
```{r, include=FALSE}
# Data preprocessing, select and calculate the required columns
selected_data_1 <- crimedata %>%
filter(state == "NJ") %>% # Filter only New Jersey
select(state, communityName, population, murders, rapes, robberies, assaults, burglaries, larcenies, autoTheft, arsons, perCapInc, PctLess9thGrade, PctNotHSGrad, PctBSorMore) %>%
mutate(
murders = ifelse(is.na(murders), median(murders, na.rm = TRUE), murders),
rapes = ifelse(is.na(rapes), median(rapes, na.rm = TRUE), rapes),
robberies = ifelse(is.na(robberies), median(robberies, na.rm = TRUE), robberies),
assaults = ifelse(is.na(assaults), median(assaults, na.rm = TRUE), assaults),
burglaries = ifelse(is.na(burglaries), median(burglaries, na.rm = TRUE), burglaries),
larcenies = ifelse(is.na(larcenies), median(larcenies, na.rm = TRUE), larcenies),
autoTheft = ifelse(is.na(autoTheft), median(autoTheft, na.rm = TRUE), autoTheft),
arsons = ifelse(is.na(arsons), median(arsons, na.rm = TRUE), arsons)
) %>%
mutate(
total_crime = murders + rapes + robberies + assaults + burglaries + larcenies + autoTheft + arsons,
crime_rate = total_crime / population * 100000
) %>%
select(-murders, -rapes, -robberies, -assaults, -burglaries, -larcenies, -autoTheft, -arsons)
selected_data_1 <- selected_data_1 %>%
mutate(
NoHigherEdu = PctLess9thGrade + PctNotHSGrad,
HigherEdu = PctBSorMore
) %>%
select(state, communityName, population, perCapInc, NoHigherEdu, HigherEdu, total_crime, crime_rate)
# Convert to long format data
long_data_1 <- selected_data_1 %>%
pivot_longer(
cols = c(NoHigherEdu, HigherEdu),
names_to = "Education_Level",
values_to = "Percentage"
)
```
### Crime Rate vs Per Capita Income
```{r}
# Plot the crime rate versus income per capita
p1 <- ggplot(selected_data_1, aes(x = perCapInc, y = crime_rate)) +
geom_point(color = "#69b3a2", size = 1.5) +
geom_smooth(method = "lm", formula = y ~ x, se = FALSE, color = "#a89181", linetype = "dashed") +
ggtitle("Crime Rate vs Per Capita Income (NJ)") +
labs(y = "Crime Rate", x = "Per Capita Income") +
theme_minimal()
ggplotly(p1)
```
### Crime Rate vs Education Level
```{r}
custom_colors <- c("#779de9", "#f8c1b7")
# Plot the crime rate versus education level
p2 <- ggplot(long_data_1, aes(x = Percentage, y = crime_rate, color = Education_Level, shape = Education_Level)) +
geom_point(alpha = 0.8, size = 2, aes(text = paste("Education Level:", Education_Level, "<br>Proportion:", round(Percentage, 2), "%"))) +
geom_smooth(method = "lm", formula = y ~ x, se = FALSE) +
labs(
title = "Crime Rate vs Education Level (NJ)",
x = "Proportion of Education Level (%)",
y = "Crime Rate",
shape = "Education Level",
color = "Education Level"
) +
scale_color_manual(values = custom_colors) +
theme_minimal()
ggplotly(p2, tooltip = "text") %>% layout(showlegend = TRUE)
```
Relationship between age and violent and nonviolent crime
==============================
Column{data-width=200}
-------------------------------
```{r}
library(ggplot2)
library(tidyverse)
library(plotly)
data <- read.csv("crimedata.csv")
```
```{r}
data_long <- data %>%
select(agePct12t21, agePct12t29, agePct16t24, agePct65up, ViolentCrimesPerPop, nonViolPerPop) %>%
pivot_longer(cols = starts_with("agePct"), names_to = "Age_Group", values_to = "Percentage")
```
```{r}
lm_model <- lm(ViolentCrimesPerPop ~ Percentage + Age_Group, data = data_long)
data_long <- data_long %>%
mutate(predicted = predict(lm_model, newdata = data_long))
p1 <- plot_ly(data_long,
x = ~Percentage,
y = ~ViolentCrimesPerPop,
color = ~Age_Group,
text = ~paste("Age Group:", Age_Group, "<br>Violent Crimes:", ViolentCrimesPerPop),
type = "scatter",
mode = "markers") %>%
add_trace(x = ~Percentage,
y = ~predicted,
mode = 'lines',
line = list(color = 'Age Group'),
name = 'Regression Line') %>%
layout(title = "Age Percentage vs Violent Crimes per Population",
xaxis = list(title = "Age Percentage"),
yaxis = list(title = "Violent Crimes per Population"))
p1
```
```{r}
lm_non_violent <- lm(nonViolPerPop ~ Percentage + Age_Group, data = data_long)
data_long <- data_long %>%
mutate(predicted_non_violent = predict(lm_non_violent, newdata = data_long))
p2 <- plot_ly(data_long,
x = ~Percentage,
y = ~nonViolPerPop,
color = ~Age_Group,
text = ~paste("Age Group:", Age_Group, "<br>Non-Violent Crimes:", nonViolPerPop),
type = "scatter",
mode = "markers") %>%
add_trace(x = ~Percentage,
y = ~predicted_non_violent,
mode = 'lines',
line = list(color = 'Age Group'),
name = 'Regression Line') %>%
layout(title = "Age Percentage vs Non-Violent Crimes per Population",
xaxis = list(title = "Age Percentage"),
yaxis = list(title = "Non-Violent Crimes per Population"))
p2
```
```{r}
data <- read.csv("crimedata.csv")
```
Race and Population
=====================================
Row
-------------------------------
### race vs ViolentCrimesPerPop
```{r}
data <- data %>%
mutate(
dominant_race = case_when(
racepctblack == pmax(racepctblack, racePctWhite, racePctAsian, racePctHisp) ~ "Black",
racePctWhite == pmax(racepctblack, racePctWhite, racePctAsian, racePctHisp) ~ "White",
racePctAsian == pmax(racepctblack, racePctWhite, racePctAsian, racePctHisp) ~ "Asian",
racePctHisp == pmax(racepctblack, racePctWhite, racePctAsian, racePctHisp) ~ "Hispanic",
TRUE ~ "Other"
)
)
p3 <- plot_ly(data,
x = ~state,
y = ~ViolentCrimesPerPop,
text = paste("community:", data$communityName,
"ViolentCrimesPerPop:", data$ViolentCrimesPerPop,
"Race:", data$dominant_race),
type = "scatter",
mode = "markers",
color = ~dominant_race,
colors = c("Black" = "#000000",
"White" = "#ffc0cb",
"Asian" = "#ffff00",
"Hispanic" = "#FF6347"),
marker = list(size = 10,opacity = 0.7)
) %>%
layout(xaxis = list(title="State"),
yaxis = list(title = "ViolentCrimesPerPop"),
showlegend = TRUE)
p3
```
Row
-------------------------------
### race vs nonViolPerPop
```{r}
data <- data %>%
mutate(
dominant_race = case_when(
racepctblack == pmax(racepctblack, racePctWhite, racePctAsian, racePctHisp) ~ "Black",
racePctWhite == pmax(racepctblack, racePctWhite, racePctAsian, racePctHisp) ~ "White",
racePctAsian == pmax(racepctblack, racePctWhite, racePctAsian, racePctHisp) ~ "Asian",
racePctHisp == pmax(racepctblack, racePctWhite, racePctAsian, racePctHisp) ~ "Hispanic",
TRUE ~ "Other"
)
)
p3 <- plot_ly(data,
x = ~state,
y = ~nonViolPerPop,
text = paste("community:", data$communityName,
"nonViolPerPop:", data$nonViolPerPop,
"Race:", data$dominant_race),
type = "scatter",
mode = "markers",
color = ~dominant_race,
colors = c("Black" = "#000000",
"White" = "#ffc0cb",
"Asian" = "#ffff00",
"Hispanic" = "#FF6347"),
marker = list(size = 10,opacity = 0.7)
) %>%
layout(xaxis = list(title="State"),
yaxis = list(title = "nonViolPerPop"),
showlegend = TRUE)
p3
```
Row
-------------------------------
### population vs race
```{r}
library(tidyr)
data <- data %>%
select(state, racepctblack, racePctWhite, racePctAsian, racePctHisp,communityName) %>%
gather(key = "race", value = "percentage", racepctblack, racePctWhite, racePctAsian, racePctHisp) %>%
mutate(race = factor(race, levels = c("racepctblack", "racePctWhite", "racePctAsian", "racePctHisp"),
labels = c("Black", "White", "Asian", "Hispanic")),
population_count = data$population * percentage / 100)
p31 <- plot_ly(data,
x = ~state,
y = ~population_count,
text = paste("community:", data$communityName,
"population:", data$population,
"Race:", data$race),
type = "bar",
mode = "markers",
color = ~race,
colors = c("Black" = "#000000",
"White" = "#ffc0cb",
"Asian" = "#ffff00",
"Hispanic" = "#FF6347"),
marker = list(size = 10,opacity = 0.7)
) %>%
layout(barmode = "stack",
xaxis = list(title="State"),
yaxis = list(title = "Population"),
showlegend = TRUE)
p31
```
Data Table
========================================
```{r}
cdata <- read.csv("processed_crimedata.csv")
```
```{r}
datatable(cdata,
caption = "Failure Data",
rownames = T,
filter = "top",
options = list(pageLength = 25))
```
Summary
========================================
These analyses reveal a complex relationship between crime rates and economic factors, racial composition and age, suggesting that crime is not just a single social problem, but is inffuenced by the interaction of multiple factors. An in-depth analysis of the association between a number of social factors and crime rates in the United States reveals that areas with higher per capita incomes typically have lower crime rates, while areas with higher levels of poverty and lower levels of education tend to face higher crime pressures. This phenomenon may be closely related to the scarcity of resources, higher unemployment and the existence of social inequalities.
In addition, racial composition plays a signiffcant role in crime rates. There are differences in the extent to which different racial groups are involved in different types of crime, and certain racial groups may be exposed to a higher risk of crime in particular neighborhoods.
These analyses provide a more complete understanding of the spatial distribution of crime rates across U.S. states and the multiple factors that inffuence this distribution. This understanding provides policymakers with a more speciffc social context to help them design more targeted and effective
interventions to reduce crime rates.